After cerebral ischemia or trauma, microglia are initially activated into a damaging inflammatory phenotype. But eventually, they may switch to a beneficial pro-remodeling phenotype as the brain tries to recover. This project dissects a novel mechanism that allows cerebral endothelium and reactive astrocytes todifferentially regulate the microglial switch. Our hypothesis is as follows. Normally, brain microglia cannot see cerebral endothelium because they reside behind the BBB. During acute injury, the BBB is leaky and microglia now see microparticles that are released by damaged or inflammed cerebral endothelium. This activates them into a deleterious phenotype (e.g. high TNF? and IL1?). After this initial injury stage subsides, the blood-brain barrier (BBB) is repaired and reactive astrocytes emerge that begin to release pro-recovery mediators such as high-mobility-groupbox-1 (HMGB1), IL-4 and IL-10. This causes microglia to switch to a beneficial phenotype (e.g. high growth factors). Our project will dissect the inter- and intra-cellular mechanisms involved in this gliovascular regulation of the microglial switch. In Aim 1, we use cell cultures to show that conditioned media from endothelium vs astrocytes have opposite effects on microglia activation. Then we show that harmful vs beneficial microglia have either neurotoxic or neuroplastic effects when added to primary neurons. In Aim 2, we will confirm that the differential effects of endothelium vs astrocytes is mediated by release of microparticles, HMGB1, IL-4 and IL-10. In Aim 3, we will correlate the temporal profiles of BBB leakage, astrocytosis and biphasic microglia after focal cerebral ischemia or trauma in mice. We will also directly inject endothelial microparticles oractivate astrocytes via optogenetics to see whether we can evoke the predicted microglial phenotypes in vivo. Synergizing with Project 2, we will try to asses these microglial pathways in white matter as well. And synergizing with Project 3, we will try to manipulate rho-kinase pathways to switch microglia into a beneficial mode and promote recovery. Imaging and optogenetics will be supported by Core A. Neurorecovery in our mouse models will be supported by Core B. Taken together, our studies should define a mechanism by which endothelium and astrocytes regulate the microglial switch in the remodeling neurovascular unit. These findings may eventually lead to therapeutic opportunities for promoting repair and recovery after stroke and neurodegeneration.